Stem Cell Reports

Article

Activin Is Superior to BMP7 for Efficient Maintenance of Human iPSC-Derived

Nephron Progenitors
Shunsuke Tanigawa,1 Hidekazu Naganuma,1,2 Yusuke Kaku,1 Takumi Era,3 Tetsushi Sakuma,4
Takashi Yamamoto,4 Atsuhiro Taguchi,1,5 and Ryuichi Nishinakamura1,*
1Department of Kidney Development, Institute of Molecular Embryology and Genetics, Kumamoto University, Kumamoto 860-0811, Japan
2Department of Urology, Kyushu University Graduate School of Medical Science, Fukuoka 812-8582, Japan
3Department of Cell Modulation, Institute of Molecular Embryology and Genetics, Kumamoto University, Kumamoto 860-0811, Japan
4Department of Mathematical and Life Sciences, Graduate School of Science, Hiroshima University, Hiroshima 739-8526, Japan
5Present address: Department of Genome Regulation, Max Planck Institute for Molecular Genetics, 14195 Berlin, Germany

*Correspondence: ryuichi@kumamoto-u.ac.jp
https://doi.org/10.1016/j.stemcr.2019.07.003

SUMMARY

Kidney formation is regulated by the balance between maintenance and differentiation of nephron progenitor cells (NPCs). Now that
directed differentiation of NPCs from human induced pluripotent stem cells (iPSCs) can be achieved, maintenance and propagation
of NPCs in vitro should be beneficial for regenerative medicine. Although WNT and FGF signals were previously shown to be essential
for NPC propagation, the requirement for BMP/TGFb signaling remains controversial. Here we reveal that activin has superior effects
to BMP7 on maintenance efficiency of human iPSC-derived NPCs. Activin expanded ITGA8+/PDGFRA
/SIX2-GFP+ NPCs by 5-fold per
week at 80%–90% efficiency, and the propagated cells possessed robust capacity for nephron formation both in vitro and in vivo. The
expanded cells also maintained their nephron-forming potential after freezing. Furthermore, the protocol was applicable to multiple
non-GFP-tagged iPSC lines. Thus, our activin-based protocol will be applicable to a variety of research fields including disease modeling
and drug screening.

INTRODUCTION these goals. Indeed, several groups, including ours, have

The kidney contains a large number of nephrons, which
are functional units consisting of glomeruli and renal tu-
bules. The kidney develops by interactions between the
metanephric mesenchyme and the ureteric bud, the former
of which contains nephron progenitor cells (NPCs) that
form the epithelia of glomeruli and renal tubules in
response to wingless-type mouse mammary tumor virus
integration site family member (WNT) signals emitted
from the latter. At the same time, the ureteric bud-derived
signals maintain and propagate some NPCs in the undiffer-
entiated state, and the balance between NPC maintenance/
propagation and differentiation is critical for kidney devel-
opment (Costantini and Kopan, 2010). NPCs cease propa-
gation and become terminally differentiated within a few
days after birth in mice (Hartman et al., 2007) and at
approximately 36 weeks of gestation in humans (Lind-
strom et al., 2018a).
In recent years, we and others have succeeded in NPC
induction and resultant kidney organoid generation
from mouse embryonic stem cells (ESCs) and/or human
induced pluripotent stem cells (iPSCs) (Morizane et al.,
2015; Taguchi et al., 2014; Taguchi and Nishinakamura,
2017; Takasato et al., 2016). Thus, NPC purification and
expansion should have an impact on regenerative
medicine targeting cell therapy, disease modeling, drug
screening, and transplantable kidney reconstruction,
because large numbers of NPCs are required to achieve
reported expansion protocols for NPCs derived from
ESCs or iPSCs (Brown et al., 2015; Li et al., 2016; Tanigawa
et al., 2015, 2016), all of which involved fibroblast growth
factor (FGF) and WNT signals, consistent with previous
findings in vivo (Barak et al., 2012). However, the utiliza-
tion of bone morphogenetic protein (BMP) signaling mol-
ecules and their inhibitors differed depending on the
protocol. Brown et al. (2015) reported expansion of hu-
man iPSC-derived NPCs by addition of BMP7 and BMP re-
ceptor inhibitor LDN193189. However, their NPCs only
formed renal tubule-like structures and not glomeruli.
When the same protocol was applied to NPCs from hu-
man embryos, some NPC genes were maintained, but
the expanded cells failed to show nephron formation his-
tologically (Pode-Shakked et al., 2017). We reported
expansion of NPCs from mouse embryos by adding leuke-
mia inhibitory factor (LIF) and low concentrations of
BMP7. This protocol was applicable to NPCs derived
from mouse ESCs and human iPSCs, and generated both
glomeruli and renal tubules, emphasizing that the full po-
tency for nephron formation was maintained in the
expanded NPCs. However, the expansion period for hu-
man iPSC-derived NPCs was limited (1–2 weeks). Subse-
quently, Li et al. (2016) reported long-term expansion of
NPCs from mouse and human embryos. They utilized a
spheroid culture in the presence of LIF, BMP7, and combi-
nation of BMP and transforming growth factor b (TGFb)
receptor inhibitors, and obtained propagation of human

322 Stem Cell Reports j Vol. 13 j 322–337 j August 13, 2019 j ª 2019 The Author(s).
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Figure 1. NPCs Are Enriched in the ITGA8+/PDGFRA
/SIX2-GFP+ Fraction in Human iPSC-Derived Kidney Tissues
(A) Generation of SIX2-GFP iPSCs by TALENs. A GFP cassette was inserted into exon 1 in the SIX2 gene by homologous recombination. The
selection cassette was removed by Cre-mediated excision. Southern blot analysis of wild-type (WT) (+/+), heterozygous (+/
), and
homozygous (
/
) clones (right). The positions of the probe for Southern blotting and PCR primers are indicated on the map (left).
(B) Isolation of nephron progenitors from the ITGA8+/PDGFRA
/SIX2-GFP+ fraction in induced NPCs. The gates in the dot plots show SIX2-
GFP+ cells in the total cells (left), ITGA8+/PDGFRA
cells in the total cells (center), and ITGA8+/SIX2-GFP+ cells in the ITGA8+/PDGFRA

fraction (right). I, ITGA8; P, PDGFRA; SIX2+, SIX2-GFP+. The percentage of ITGA8+/PDGFRA
/SIX2-GFP+ cells in the total cells is shown on
the right (I+/P
/SIX2+ cells/total cells). Data represent means ± SEM from six independent experiments.
(legend continued on next page)

Stem Cell Reports j Vol. 13 j 322–337 j August 13, 2019 323

embryo-derived NPCs for 7 months. They further claimed
that the same protocol was applicable to human iPSC-
derived NPCs. However, none of these reports addressed
the quantitative percentages of human NPCs. If NPCs
constitute only a minor population of the expanded cells,
nephrogenesis will occur in limited areas of the tissues,
and thus hamper reproducible disease modeling or drug
screening. Thus, it is necessary to establish methods that
can measure the precise percentages of functional NPCs
in expanded cells.
The transcription factor SIX2 is specifically expressed in
NPCs and antagonizes their Wnt-mediated differentiation
(Kobayashi et al., 2008). Li et al. (2016) generated SIX2-
GFP iPSCs, but did not perform quantitative analysis of
GFP+ NPCs. Meanwhile, we reported that NPCs from
mouse embryos and human iPSCs could be sorted as an
ITGA8+/PDGFRA
population (Kaku et al., 2017; Taguchi
et al., 2014). ITGA8 was expressed in NPCs from both
mice and humans in vivo (O’Brien et al., 2016), and deletion
of Itga8, as well as its ligand nephronectin, led to kidney hy-
poplasia in mice (Muller et al., 1997), indicating that ITGA8
expression in NPCs is critical for kidney development. Sin-
gle-cell RNA-sequencing (RNA-seq) analysis of nephron
progenitors in human embryos further revealed that they
were ITGA8+/PDGFRA
(Lindstrom et al., 2018b; O’Brien
et al., 2016). Therefore, we reasoned that the combination
of these two criteria (SIX2-GFP+ and ITGA8+/PDGFRA
)
would lead to more reliable and quantitative readouts for
measurement of maintenance efficiency of NPCs upon cul-
ture. For this purpose, we generated SIX2-GFP human
iPSCs and measured the percentages of ITGA8+/
PDGFRA
/SIX2-GFP+ NPCs, leading to the unexpected
finding that activin A (activin) has superior effects to
BMP7 on the maintenance efficiency of human iPSC-
derived NPCs.
structed a pair of plasmids expressing TALENs that tar-
geted regions in close proximity to the SIX2 start codon,
and introduced these plasmids together with a targeting
vector containing the GFP gene and homology arms into
iPSCs (Figure 1A). We successfully obtained heterozygous
GFP-knockin clones and induced them toward NPCs,
based on our previously published protocol (Taguchi
et al., 2014; Taguchi and Nishinakamura, 2017). At day
13, SIX2-GFP+ cells constituted 7.2% of the induced
tissues (n = 6; Figure 1B). Meanwhile, the ITGA8+/
PDGFRA
population represented 22.7%, and 18.1%
of these cells were SIX2-GFP+. Thus, 4.1% of the total
cells were ITGA8+/PDGFRA
/SIX2-GFP+. Expression of
NPC marker Eya1 in ITGA8+/PDGFRA
/SIX2-GFP+ or
ITGA8+/PDGFRA
/SIX2-GFP
cells was higher than that
in ITGA8
/PDGFRA
/SIX2-GFP
cells, while the expres-
sion levels of other NPC markers, such as SIX2,
MEOX1, and HOXA11, were higher in ITGA8+/
PDGFRA
/SIX2-GFP+ cells even when compared with
ITGA8+/PDGFRA
/SIX2-GFP
cells (Figure 1C). When
these fractions were aggregated and combined with em-
bryonic spinal cord, a classical inducer of nephrogenesis,
ITGA8+/PDGFRA
/SIX2-GFP+ cells formed nephron
epithelia of glomeruli and renal tubules (CADHERIN6
[CDH6]: proximal tubules; NEPHRIN: podocytes;
E-CADHERIN [CDH1]: distal tubules) that constituted
53.3% of the induced kidney tissues (Figures 1D and
1E). In contrast, ITGA8+/PDGFRA
/SIX2-GFP
cells ex-
hibited minimal nephrogenesis (8.7% of the whole tis-
sues). The ITGA8
/PDGFRA
/SIX2-GFP
fraction also
showed minimal nephrogenesis (Figure S1), consistent
with our previous report (Kaku et al., 2017). Thus, hu-
man iPSC-derived NPCs were preferentially populated
in the ITGA8+/PDGFRA
/SIX2-GFP+ fraction.

BMP7 Is Insufficient to Maintain Human iPSC-Derived

RESULTS
Human iPSC-Derived NPCs Are Enriched in the
ITGA8+/PDGFRA
/SIX2-GFP+ Fraction
To isolate NPCs, we inserted GFP into the SIX2 locus of
human iPSCs by homologous recombination. We con-
NPCs
To propagate iPSC-derived ITGA8+/PDGFRA
/SIX2-GFP+
cells, we aggregated the sorted cells and cultured them as
spheres for 7 days under our previously defined conditions,
which included FGF9, CHIR99021 (CHIR), LIF, DAPT
(N-[N-(3,5-difluorophenacetyl)-L-alanyl]-S-phenylglycine

(C) Progenitor marker expression in each progenitor fraction. Isolated cells from the indicated fractions were analyzed for gene expression
by qPCR. Data represent means ± SEM from three independent experiments. *p < 0.05, **p < 0.01 versus ITGA8
/PDGFRA
/SIX2-GFP
; NS,
not significant.
(D) Histology and immunostaining of spinal cord-induced tissues from isolated NPCs at day 9 after induction. Magnified images of the
ITGA8+/PDGFRA
/SIX2-GFP+ fraction are shown on the right. CDH6 (green) shows proximal tubules, NEPHRIN (red) shows glomeruli, and
CDH1 (yellow) shows distal tubules. Scale bars: 200 mm (left and middle) and 50 mm (right).
(E) Quantification of nephron areas in induced tissues. The percentages of nephron structures in the induced tissues relative to the total
section areas were calculated and shown as bar graphs. **p < 0.01 versus ITGA8+/PDGFRA
/SIX2-GFP
(n = 3 per group).
See also Figure S1.

324 Stem Cell Reports j Vol. 13 j 322–337 j August 13, 2019


t-butyl ester), and BMP7 (Figure 2A) (Tanigawa et al., 2016).
We previously proposed that a relatively low concentration
of BMP7 (5 ng/mL) was beneficial for maintenance
of NPCs, while other reports utilized higher concentrations
(30–50 ng/mL) along with BMP receptor inhibitors (Barak
et al., 2012; Brown et al., 2015; Li et al., 2016). Thus,
we examined the effects of two concentrations of BMP7
(5 ng/mL versus 50 ng/mL). However, the percentages
of SIX2-GFP+ cells were 70.7% and 58.3%, respectively,
and ITGA8+/PDGFRA
/SIX2-GFP+ cells constituted only
Figure 2. BMP7 Is Insufficient to Maintain
Human iPSC-Derived NPCs
(A) Schematic diagram of the culture pro-
cedure for purified NPCs. Purified NPCs were
aggregated and cultured in basal medium
containing FGF9, CHIR, LIF, DAPT, and BMP7
(5 or 50 ng/mL) for 7 days.
(B) Flow-cytometry quantification of
ITGA8+/PDGFRA
/SIX2-GFP+ (I+/P
/SIX2+)
cells after culture with BMP7 (5 ng/mL) for
7 days. The percentage of ITGA8+/PDGFRA
/
SIX2-GFP+ cells in the total cells is shown on
the right (I+/P
/SIX2+ cells/total cells).
n = 3 per group.
(C) Flow-cytometry quantification of ITGA8+/
PDGFRA
/SIX2-GFP+ cells after culture with
BMP7 (50 ng/mL) for 7 days (n = 3 per
group).
(D) Numbers of total and ITGA8+/PDGFRA
/
SIX2-GFP+ cells after 7 days of culture (n = 3
per group).
(E) qPCR analysis of progenitor markers in
NPCs after culture for 7 days. Freshly isolated
ITGA8+/PDGFRA
/SIX2-GFP+ cells were used
as a positive control (Day 0). Data represent
means ± SEM from three independent ex-
periments. *p < 0.05, **p < 0.01, cells
cultured with BMP7 versus Day 0.
See also Figure S2.
24.1% and 26.9% of the total cells, respectively (Figures
2B and 2C). Although the cell numbers increased signifi-
cantly during the 7-day culture, the expansion of the
ITGA8+/SIX2-GFP+ fraction was only 2.7-fold and 1.9-
fold, respectively (Figure 2D). Consistent with these obser-
vations, the expression levels of NPC markers in the whole
spheres decreased while those of differentiation markers
(LHX1, CDH1) increased (Figure 2E). Thus, BMP7-based
conditions were insufficient for maintenance of ITGA8+/
PDGFRA
/SIX2-GFP+ NPCs in vitro, although the lower
Stem Cell Reports j Vol. 13 j 322–337 j August 13, 2019 325
Figure 3. Activin Maintains NPCs at High Efficiency
(A) Quantification of ITGA8+/PDGFRA
/SIX2-GFP+ (I+/P
/SIX2+) cells cultured for 7 days with BMP7 or TGFb-related ligands. NPCs were
cultured for 7 days in basal medium (as shown in Figure 2A) with addition of BMP7 (50 ng/mL) or TGFb-related ligands activin (50 ng/mL),
GDF11 (50 ng/mL), TGFb1 (50 ng/mL), or TGFb2 (50 ng/mL), then analyzed by flow cytometry. The percentage of ITGA8+/PDGFRA
/SIX2-
GFP+ cells in the total cells is shown on the right (I+/P
/SIX2+ cells/total cells). n = 3 per group.
(B) Numbers of total and ITGA8+/PDGFRA
/SIX2-GFP+ cells after culture for 7 days (n = 3 per group).
(C) qPCR analysis of NPC markers. Freshly isolated ITGA8+/PDGFRA
/SIX2-GFP+ cells were used as a positive control (Day 0). Data represent
means ± SEM from three independent experiments. **p < 0.01 versus day 0.
(D) Reconstitution of three-dimensional nephron structures by NPCs cultured with BMP7 or activin for 7 days. H&E-stained images (upper)
and immunostained images (lower) of induced nephrons from NPCs cultured with BMP7 or activin are shown. Scale bars, 200 mm.
(E) Quantification of nephron areas in induced tissues. The percentages of nephron-like structures in the induced tissues related the total
section areas were calculated and shown as bar graphs. Data represent means ± SEM from three independent experiments. **p < 0.01, cells
cultured with BMP7 versus activin.
(F) Magnified images of glomerular podocytes generated from NPCs expanded with activin. COL4, type IV collagen. Arrowheads indicate
pre-slit diaphragm domains; arrows indicate basal pre-slit diaphragm domains. Scale bar, 10 mm.
(G) H&E staining of kidney tissues generated from thawed NPCs at day 20 after spinal cord induction. Scale bar, 200 mm.

(legend continued on next page)

326 Stem Cell Reports j Vol. 13 j 322–337 j August 13, 2019

BMP7 concentration showed slightly better degrees of
propagation and NPC marker retention.
Li et al. (2016) reported that the combination of BMP7,
BMP receptor inhibitor LDN193189, and TGFb receptor in-
hibitor A83-01 was effective for 7-month in vitro expansion
of NPCs derived from human embryos. When we applied
their conditions to our iPSC-derived ITGA8+/PDGFRA
/
SIX2-GFP+ cells, robust cell proliferation was observed
and 71.5% of the cells remained SIX2-GFP+ (Figures S2A
and S2B). However, the percentages of the ITGA8+/
PDGFRA
fraction were low, even at day 7 of culture
(25.5%), and only 18.2% of the expanded cells remained
ITGA8+/PDGFRA
/SIX2-GFP+. RT-PCR analysis showed
that SIX2 expression was relatively maintained, consistent
with the high SIX2-GFP+ cell percentage, but Eya1 was
reduced and a differentiation marker, CDH1, was increased
at day 7 (Figure S2C). Therefore, monitoring of SIX2 expres-
sion alone could lead to over-estimation of the mainte-
nance efficiency of NPCs.
Activin Maintains NPCs at High Efficiency
Among the various reagents tested for improvement of the
culture condition, such as BMP4, FGF10, FGF20, CXCL12,
Wnt9b, and R-spondin, we found that TGFb-related li-
gands, including activin, GDF11, TGFb1, and TGFb2, led
to better maintenance of ITGA8+/PDGFRA
/SIX2-GFP+
cells than BMP7 at day 7 of culture (Figure 3A). Among
them, activin was the most effective: 91.6% of cells were
SIX2-GFP+ and up to 83.1% of cells remained ITGA8+/
PDGFRA
/SIX2-GFP+ (Figures 3A and S3). Furthermore, ac-
tivin achieved 5.3 ± 0.8-fold expansion (n = 3) of NPCs,
which was superior to BMP7 (Figure 3B). The expression
levels of SIX2 and other progenitor markers, as well as dif-
ferentiation markers, were comparable with those at the
starting point (day 0), in sharp contrast to those with
BMP7 treatment (Figure 3C). Comprehensive RNA-seq
and clustering analyses also placed the NPCs cultured
with activin closer to those at the start of culture and in
human embryos, compared with the NPCs cultured with
BMP7 (Figure 4). The activin-based protocol maintained
nephron progenitor markers, including SIX2, SALL1,
EYA1, CITED1, PAX2, WT1, MEOX1, and ITGA8. Human
NPC-specific genes, such as LYPD1, TMEM100, WASF3,
UNC5B, and PHF19, which were recently identified in hu-
man embryos (Lindstrom et al., 2018b), were also main-
tained. In contrast, SIX2 and MEOX1 were significantly
downregulated upon BMP7 treatment while differentia-
(H) Immunostaining of nephrons generated from thawed NPCs. CDH6 (green), NEPHRIN (red), and CDH1 (yellow) are shown. Scale bar,
200 mm.
(I) Magnified image of glomerular podocytes. Immunostaining for NEPHRIN (red) and COL4 (green) are shown. Arrowheads indicate pre-
slit diaphragm domains; arrows indicate basal pre-slit diaphragm domains. Scale bar, 10 mm.
See also Figures S3 and S4.
tion markers (LHX1, CDH1, HNF1B, JAG1, GATA3, and
EMX2) were upregulated. Therefore, activin can maintain
human iPSC-derived ITGA8+/PDGFRA
/SIX2-GFP+ NPCs
more effectively than BMP7. Next, we examined the
nephron-forming potential of the expanded cells. Cells
expanded with activin or other TGFb-related ligands ex-
hibited robust formation of glomeruli and renal tubules
at day 20 after spinal cord induction, and the areas of
nephrogenesis were much larger than that for cells treated
with BMP7 (Figures 3D and 3E, n = 6; Figure S4). Among
them, activin showed the largest area of nephrogenesis
(Figure S4, n = 3), and was thus selected for use in subse-
quent experiments. Notably, NEPHRIN+ rod or dot-like
structures were observed in the glomerular podocytes
generated from activin-expanded NPCs. These structures
were detected on the lateral domains of the podocytes,
as well as on their basal domains adjacent to the type IV
collagen (COL4)-positive basement membrane (Figure 3F).
We recently reported that these structures observed in
iPSC-derived kidney organoids represent the precursors
of slit diaphragms, which play a critical role in filtration
of blood to produce urine (Tanigawa et al., 2018). Thus,
the activin-based protocol can retain NPCs at comparable
quality to the start of the culture for the generation of
these nephron structures.
We further examined whether the expanded NPCs could
be frozen before induction to form nephrons. NPCs were
cultured for 7 days with activin, frozen in single-cell sus-
pensions, and thawed for regeneration of NPC spheres.
When the spheres were induced with spinal cord they
readily exhibited nephrogenesis, including glomerular po-
docytes with slit diaphragm precursors (Figures 3G–3I). The
nephron areas constituted 79.8% ± 3.6% (n = 4), and were
comparable with those of unfrozen progenitors. These re-
sults suggest that the complicated iPSC handling required
for NPC induction can be bypassed by starting with frozen
NPC stocks.
Each of the Added Factors Is Required for Proper NPC
Expansion
To confirm the requirement of each added factor for NPC
maintenance and propagation, we excluded each factor
from the optimized medium and performed assessment
on day 7 of culture (Figure 5A). When activin was elimi-
nated, the percentage of the ITGA8+/PDGFRA
/SIX2-
GFP+ fraction decreased dramatically to 42% as expected,
while the non-progenitor cell numbers were increased
Stem Cell Reports j Vol. 13 j 322–337 j August 13, 2019 327

(Figures 5A–5C). Depletion of DAPT significantly reduced
the percentage of the progenitor fraction, as well as the
numbers of progenitor cells. FGF9 elimination resulted in
a sharp drop in the total cell numbers, indicating that
this factor is essential for the proliferation of both progen-
itors and non-progenitors, consistent with our previous
328 Stem Cell Reports j Vol. 13 j 322–337 j August 13, 2019
Figure 4. RNA-Seq Analysis of the Propa-
gated NPCs
Heatmap and gene clustering of NPCs under
the conditions listed below. Human NPCs
from a 16- week embryonic kidney were used
as a positive control (Lindstrom et al.,
2018b). I+/P
/SIX2+, ITGA8+/PDGFRA
/
SIX2-GFP+ cells before expansion; I+/P
/
SIX2
, ITGA8+/PDGFRA
/SIX2-GFP
cells
before expansion; h16wk SIX2+/MEIS1+,
hMARIS_SIX2+_MEIS1+ cells from a human
16-week embryonic kidney (Lindstrom et al.,
2018b); h16wk SIX2+, hMARIS_SIX2+ cells
from a human 16-week embryonic kidney
(Lindstrom et al., 2018b); h16wk ITGA8+,
ITGA8+ cells from a human 17-week embry-
onic kidney (O’Brien et al., 2016); ActivinA
Day 7, ITGA8+/PDGFRA
/SIX2-GFP+ cells
cultured with activin for 7 days; BMP7 Day 7,
ITGA8+/PDGFRA
/SIX2-GFP+ cells cultured
with BMP7 for 7 days. NPC markers (Park
et al., 2012) are shown in red. Human-spe-
cific NPC markers (Lindstrom et al., 2018b)
are shown in blue.
See also Table S1.
findings in mouse progenitor culture (Tanigawa et al.,
2016). Elimination of LIF or CHIR did not affect the per-
centage of the progenitor fraction (Figures 5A and 5B),
but the numbers of progenitor cells were significantly
decreased (Figure 5C). Thus, each factor is necessary for
proper NPC maintenance and propagation.
Figure 5. Activin-Mediated Signal Is Essential for Maintenance of NPCs
(A) Quantification of ITGA8+/PDGFRA
/SIX2-GFP+ (I+/P
/SIX2+) cells after culture for 7 days. NPCs were cultured for 7 days in the
optimized medium including activin (50 ng/mL) as indicated (All), and with individual factors eliminated from the optimized medium,
then analyzed by flow cytometry. The percentages of ITGA8+/PDGFRA
/SIX2-GFP+ cells in the total cells are shown on the right (I+/P
/
SIX2+ cells/total cells) (n = 3 per group). *p < 0.05, **p < 0.01 versus All.
(B) Percentages of ITGA8+/PDGFRA
/SIX2-GFP+ cells among the total cells (n = 3 per group). **p < 0.01 versus All.
(legend continued on next page)

Stem Cell Reports j Vol. 13 j 322–337 j August 13, 2019 329

 When the expanded cells at day 7 were passaged and
cultured for a further 7 days (day 14), NPCs increased by
25-fold and constituted 79.4% of the expanded cells (Fig-
ure 5D). Furthermore, progenitor cells retained comparable
nephron-forming potential to day 7 (Figures S5A–S5C).
However, the percentage of the ITGA8+/PDGFRA
/SIX2-
GFP+ fraction at day 21 (third generation of culture) declined
to 57.3%, although the SIX2-GFP+ fraction remained high
(91.5%; Figure S5D), indicating that further fine-tuning or
addition of an unknown signal is necessary for infinite
maintenance of nephron progenitors. We also attempted
monolayer cultures of single cells, but the percentage of
the ITGA8+/PDGFRA
/SIX2-GFP+ fraction at day 14 (second
generation of culture) declined to 37.4% (Figures S5E and
S5F), indicating that three-dimensional aggregation culture
was better than two-dimensional culture.
SMAD2/3 Phosphorylation Is Correlated with NPC
Maintenance
To elucidate the mechanisms underlying the maintenance
of NPCs, we examined the effects of BMP or TGFb recep-
tor inhibitors. When NPCs were cultured for 7 days using
the activin-based protocol in the presence of BMP recep-
tor inhibitor LDN193189 the phosphorylation levels of
SMAD1/5 were reduced, while the percentage of the
ITGA8+/PDGFRA
/SIX2-GFP+ fraction was not affected
(Figures 5E and 5F), indicating that BMP receptor-medi-
ated SMAD1/5 activity had a minimal effect on NPC
maintenance. In contrast, treatment with TGFb receptor
inhibitor A83-01 resulted in a reduction of SMAD2/3
phosphorylation and a decrease in the percentage of the
ITGA8+/PDGFRA
/SIX2-GFP+ fraction (Figures 5E and
5F). Consistent with the latter observation, SIX2 expres-
sion was reduced while E-cadherin was increased (Fig-
ure 5E). Unexpectedly, we observed increased SMAD1/5
phosphorylation upon A83-01 treatment (Figure 5E),
which may reflect the higher percentage of differentiated
cells in the samples, because SMAD1/5 activity is known
to be detectable in differentiating nephrons (Brown
et al., 2013; Fetting et al., 2014). Thus, phosphorylation
of SMAD2/3, but not SMAD1/5, is correlated with NPC
maintenance in our protocol.
It was reported that non-canonical BMP signaling mole-
cules, namely JNK and p38, play critical roles in the main-
tenance and proliferation of NPCs (Brown et al., 2011; Mu-
thukrishnan et al., 2015). However, the phosphorylation of
p38 and JNK was unaltered by LDN193189 or A83-01 treat-
ment (Figure 5E). In addition, p38 or JNK inhibitors had
minimal impacts on the NPC percentages (Figure 5G), indi-
cating that non-canonical BMP signaling is not signifi-
cantly involved in our activin-based NPC maintenance.
Thus, the canonical SMAD2/3 pathway downstream of ac-
tivin may play a major role in the maintenance of human
iPSC-derived NPCs.
Propagated NPCs Form Large Areas of Nephrons with
Vascularized Glomeruli upon Transplantation
Next, we examined the in vivo nephron-forming potential
of the expanded NPCs. We cultured NPCs with activin for
7 days, combined the cells with spinal cord for a further
2 days to initiate nephrogenesis, and transplanted the
spheres underneath the kidney capsule of immunodefi-
cient mice, as previously described (Sharmin et al., 2016).
In this setting, it is known that iPSC-derived NPCs can
form glomeruli and renal tubules, the former of which
are vascularized by host endothelial cells. Glomerular po-
docytes become more mature than those in vitro, and
NEPHRIN is expressed at the basal side of the podocytes
to form slit diaphragms, comprising molecular sieves for
filtration from the vasculature (Tanigawa et al., 2018). As
a control, we used whole spheres containing NPCs before
maintenance culture, as described in our previous report
(Sharmin et al., 2016). At day 20 after transplantation,
the control cells showed formation of glomeruli and renal
tubules (day 0 in Figure 6A). However, these nephrons
constituted only a subset (10.3%, n = 3) of the transplanted
tissues (Figures 6A and 6B). In contrast, nephrons derived
from the expanded cells occupied larger areas of the trans-
planted tissues (70.1%, n = 3), suggesting good preserva-
tion of the NPC percentage and potential. In addition,

(C) Numbers of total and ITGA8+/PDGFRA
/SIX2-GFP+ cells after culture for 7 days (n = 3 per group). *p < 0.05, **p < 0.01 versus All.
(D) Estimated numbers of total and ITGA8+/PDGFRA
/SIX2-GFP+ cells after culture for 7 or 14 days. Data represent means ± SEM from three
independent experiments.
(E) Western blotting analysis of cultured ITGA8+/PDGFRA
/SIX2-GFP+ cells after treatment with BMP or TGFb receptor inhibitors. NPCs
were cultured in activin-containing medium with or without (
) inhibitors at the indicated concentrations for 7 days. The indicated
molecules were detected by immunoblotting.
(F) Quantification of ITGA8+/PDGFRA
/SIX2-GFP+ cells after culture for 7 days. NPCs were cultured for 7 days in activin-containing medium
with or without (control) inhibitors at the indicated concentrations. The percentages of ITGA8+/PDGFRA
/SIX2-GFP+ cells were measured
by flow cytometry. Data represent means ± SEM from three independent experiments. *p < 0.05, **p < 0.01 versus control.
(G) NPCs were cultured for 7 days in activin-containing medium with p38 inhibitor SB202190 (1 mM) or JNK inhibitor SP600125 (1 mM). The
percentages of ITGA8+/PDGFRA
/SIX2-GFP+ cells were measured by flow cytometry (n = 3 per group).
See also Figure S5.

330 Stem Cell Reports j Vol. 13 j 322–337 j August 13, 2019


NEPHRIN was expressed in glomerular podocytes at their
basal side and aligned in a linear manner along the endo-
thelial cells (Figures 6C and 6D), which is characteristic of
maturation-stage podocytes equipped with basally located
slit diaphragms (Tanigawa et al., 2018). These data indicate
that the ITGA8+/PDGFRA
/SIX2-GFP+ NPCs propagated by
activin retain their nephron-forming competence in vivo
and can successfully generate maturation-stage glomerular
podocytes, further supporting the authenticity of our
protocol.
The Activin-Based Protocol Is Applicable to iPSC Lines
without a GFP Tag
Next, we examined whether our protocol was applicable to
multiple iPSC lines. We further investigated the expansion
of NPCs from iPSCs without a GFP tag, which would make
our protocol more useful. For this, we established two new
iPSC lines (RN7 and RN12) from a healthy donor using a
Figure 6. Propagated NPCs Form Large
Areas of Nephrons with Vascularized
Glomeruli upon Transplantation
(A) H&E-stained sections of tissues at
20 days after transplantation. Left panels:
transplants from iPSC-derived NPCs before
expansion culture (without sorting). Right
panels: transplants from ITGA8+/PDGFRA
/
SIX2-GFP+ cells cultured with activin for
7 days. Scale bars, 50 mm. The bottom panels
show magnified images of the squares in the
upper panels. Scale bars, 200 mm.
(B) Quantification of nephron areas in
induced tissues. The percentages of nephron
structures in the induced tissues relative to
the total section areas were calculated and
shown as bar graphs (n = 3 per group).
(C) Magnified image of a glomerulus in
transplants of ITGA8+/PDGFRA
/SIX2-GFP+
cells. Scale bar, 50 mm.
(D) Immunostaining of glomeruli. Trans-
plants of ITGA8+/PDGFRA
/SIX2-GFP+ cells
after 20 and 40 days were stained for
NEPHRIN (red) and WT1 (green) or CD31
(green). The immunostained images show
vascularized glomeruli and basolateral
localization of NEPHRIN at 20 days (left and
center) and 40 days (right) after trans-
plantation. Arrowheads indicate lateral pre-
slit diaphragm domains; arrows indicate
basal linear expression of NEPHRIN. Scale
bars, 25 mm (left); and 8 mm (center and
right).
Sendai virus vector. Both iPSC lines had normal karyotypes
and expressed pluripotent stem cell markers (Figures 7A,
7B, and S6).
First, we induced the RN7 line toward the kidney lineage,
sorted the ITGA8+/PDGFRA
cells at day 13, and cultured
them for 7 days with activin or BMP7. In the presence of ac-
tivin, cells were expanded by 11-fold and 90% of the cells
were ITGA8+/PDGFRA
(Figure 7C). In contrast, only a
4.4-fold increase was observed in the presence of BMP7,
although 84% of the expanded cells were ITGA8+/
PDGFRA
(Figure 7C). Intriguingly, staining of cytospun
cells from the ITGA8+/PDGFRA
fraction revealed that
85.1% of cells were SIX2+ in the presence of activin,
compared with only 3.4% of SIX2+ cells in the presence
of BMP7 (Figure 7D). The expression level of SIX2 was
significantly reduced in the BMP7 condition (Figure 7E).
When the cells expanded by activin were induced for
differentiation, they readily formed glomeruli and renal
Stem Cell Reports j Vol. 13 j 322–337 j August 13, 2019 331
 (legend on next page)

332 Stem Cell Reports j Vol. 13 j 322–337 j August 13, 2019

tubules, and the percentages of the nephron area were
larger than those of cells cultured with BMP7 (Figure 7F,
n = 6). Finally, the expanded cells were transplanted under
the renal capsules of immunodeficient mice as described
above. At day 20 after transplantation, the cells expanded
by activin showed robust formation of glomeruli and renal
tubules (Figures 7G and 7H, n = 4). The percentages of the
nephron area in the transplanted tissues were larger than
those of cells cultured with BMP7 (Figure 7G, n = 4). Round
and vascularized glomeruli were formed that resembled
those in transplanted human embryonic kidneys (Dekel
et al., 2003). The glomeruli contained podocytes positive
for PODOCALYXIN (PODXL), and NEPHRIN was linearly
aligned with CD31+ vascular endothelial cells (Figure 7I).
In contrast, cells expanded by BMP7 showed minimal for-
mation of nephrons upon transplantation (Figure 7G). All
of these tendencies held true for the RN12 cell line (Fig-
ure S6). Thus, activin exhibits superior effects to BMP7
for NPC expansion and nephrogenesis in vitro and in vivo,
and our protocol is applicable to multiple independent
iPSC lines as well as genetically untagged clones.
DISCUSSION
We have demonstrated that activin exerts a superior
effect to BMP7 on the maintenance of NPCs derived from
human iPSCs. The percentage of ITGA8+/PDGFRA
/SIX2-
GFP+ progenitors remained high in our activin-based pro-
tocol, and the resulting nephrons occupied the majority
of the induced tissues both in vitro and in vivo. Our
improved protocol depends on quantitative measurement
of nephron progenitors based on three markers. While
monitoring of SIX2 expression with GFP is useful, we
demonstrated that dependence on SIX2 expression alone
is insufficient and further combination with two other
markers (ITGA8+/PDGFRA
) is essential for assessment of
nephron-forming competence.
Based on these stringent criteria, we identified that
activin is superior to BMP7 in maintaining human iPSC-
derived NPCs. We and others previously reported the
importance of BMP7 for NPC propagation in both mice
and humans. Li et al. (2016) claimed that human iPSC-
derived SIX2-GFP+ cells are expandable for 2 months.
However, when we applied the same protocol, the percent-
age of ITGA8+/PDGFRA
/SIX2-GFP+ cells was low, while
that of SIX2-GFP+ cells remained relatively high, again
emphasizing the importance of the progenitor criteria.
In contrast, our activin-based protocol maintained
80%–90% of the expanded cells as ITGA8+/PDGFRA
/
SIX2-GFP+. Upon transplantation, nephron epithelia
derived from the expanded progenitors occupied large
areas of the induced tissues and glomerular podocytes
exhibited the characteristic features of those in vivo:
NEPHRIN was localized at the basal region of the podocytes
and aligned with the podocyte-endothelial boundary,
underscoring the high percentage and quality of our
expanded nephron progenitors. In addition, we demon-
strated that the expanded cells could be frozen without
losing their nephron-forming potential, suggesting that
the complicated induction processes for NPCs can be by-
passed for certain sets of experiments. Furthermore, our

Figure 7. The Activin-Based Protocol Is Applicable to iPSC Lines without a GFP Tag
(A) Normal karyotype of an established iPSC clone (RN7) from a healthy donor.
(B) Expression of stem cell markers in RN7 iPSCs. Scale bars, 200 mm.
(C) Flow-cytometry analysis of the ITGA8+/PDGFRA
fraction after a 7- day expansion culture of NPCs induced from RN7. Day 0: ITGA8+/
PDGFRA
fraction of induced NPCs before expansion culture. The percentages of the ITGA8+/PDGFRA
fraction after 7 days of culture with
activin or BMP7 are shown (n = 3 per group). Rightmost panel: numbers of total and ITGA8+/PDGFRA
(I+/P
) cells after 7 days of culture
(n = 3 per group). **p < 0.01, cells cultured with activin versus BMP7.
(D) SIX2 staining of cytospun ITGA8+/PDGFRA
cells at day 7 of expansion culture. The percentages of SIX2+ cells among the ITGA8+/
PDGFRA
cells are also shown (n = 3 per group).
(E) qPCR analysis of SIX2 expression in ITGA8+/PDGFRA
cells after culture for 7 days. Freshly isolated ITGA8+/PDGFRA
cells were used as a
positive control (Day 0). **p < 0.01, cells cultured with activin or BMP7 versus Day 0 (n = 3 per group).
(F) Immunostaining of nephron structures generated from RN7-derived ITGA8+/PDGFRA
cells expanded by activin or BMP7 for 7 days.
Scale bars: 200 mm. Right panel: percentages of nephron-like structures in the induced tissues are shown as bar graphs. Data represent
means ± SEM from three independent experiments. **p < 0.01, cells cultured with activin versus BMP7.
(G) H&E-stained sections of kidney tissues at 20 days after transplantation. Left panels: transplants from RN7-derived ITGA8+/PDGFRA

cells cultured with activin (left) or BMP7 (right) for 7 days. Scale bars, 100 mm. Rightmost panel: percentages of nephron areas in
transplanted tissues (n = 4 per group). **p < 0.01, cells cultured with activin versus BMP7.
(H) Immunostaining of nephron structures (upper panel) and H&E-staining of a glomerulus (lower panel) in the transplants from RN7-
derived ITGA8+/PDGFRA
cells cultured with activin. Scale bars, 200 mm (upper panel) and 50 mm (lower panel).
(I) Immunostaining of glomeruli in kidney tissues at 20 days after transplantation of RN7-derived NPCs. CD31 (green) and PODXL (red) or
NEPHRIN (red) are shown. Arrowheads indicate background signals of red blood cells. Scale bars, 25 mm.
See also Figure S6.

Stem Cell Reports j Vol. 13 j 322–337 j August 13, 2019 333

activin-based protocol was applicable to multiple iPSC
lines, proving its robustness.
It was an unexpected finding that activin is superior to
BMP7 for NPC maintenance. Most published maintenance
protocols are based on the importance of BMP7, especially
the non-canonical pathway downstream of BMP7, while
SMAD1 induces NPC differentiation. For example, Li
et al. (2016) reported maintenance of NPCs from human
embryos utilizing BMP7 and a BMP receptor inhibitor, re-
sulting in SMAD1 inhibition. Meanwhile, expression of ac-
tivin receptor type II or phosphorylated Smad2 in the
nephron progenitor regions was reported in mice (Rowan
et al., 2018). Furthermore, deletion of GDF11, which is ex-
pressed in nephron progenitors and the ureteric bud, led
to kidney hypoplasia (Esquela and Lee, 2003). These re-
ports may be relevant to our present finding that the acti-
vin/GDF11/TGFb-SMAD2/3 pathway is involved in NPC
maintenance in vitro. Alternatively, the responsiveness to
exogenous signals may differ between in vivo NPCs and
iPSC-derived NPCs. While further investigation is required
to address these possibilities, our data emphasize the prac-
tical usefulness of the activin-based protocol, considering
the limited access to human embryos.
However, our protocol did not achieve long-term propa-
gation. Consistent with this limitation, our RNA-seq
analysis showed incomplete preservation of the gene
expression profiles during culture (Figure 4). For example,
SIX1 and OSR1 were decreased, which may explain the
limited propagation and longevity of the nephron progen-
itors under our conditions. Among the human nephron
progenitor-specific genes that were recently identified in
human embryos, ECEL1, DAPL1, COL9A2, TNFRSF19,
and PCDH15 were reduced (Figure 4). To improve our pro-
tocol, information on the expression levels of these genes
in human embryonic NPCs at the equivalent stage to
iPSC-derived NPCs is needed for reference. RNA-seq data
at 16 weeks of gestation (Lindstrom et al., 2018b), which
were used for comparison in this study, are helpful, but
gene expression levels in nephron progenitors change
significantly over time in vivo (Chen et al., 2015). Data at
an earlier time point (4–5 weeks after fertilization), when
human nephron progenitors initially emerge, would be
needed for precise comparisons. Nonetheless, our data pro-
vide the essential gene set that can confer sufficient compe-
tence to human nephron progenitors.
Although our protocol is applicable to multiple iPSC
lines without a GFP tag, identification of additional surface
markers would improve the maintenance efficiency of
NPCs. It was reported that NPCs could be enriched
in NCAM+/CD133
/FZD7+ or NCAM+/CD133
/EpCAM

fractions from human fetal kidneys at 15–22 weeks (Har-
ari-Steinberg et al., 2013; Pode-Shakked et al., 2016,
2017), and that some NPC genes were maintained during
334 Stem Cell Reports j Vol. 13 j 322–337 j August 13, 2019
2 weeks of culture. The NGFR+/EPCAM
fraction was also
used for NPC isolation from human embryonic kidneys
at 9–14 weeks (Li et al., 2016). However, our RNA-seq anal-
ysis showed that the expression of NCAM, FZD7, and NGFR
were relatively lower than that of ITGA8 in iPSC-derived
NPCs (ITGA8+/PDGFRA
/SIX2-GFP+) before expansion,
while the same genes were expressed to some extent in
SIX2+ NPCs in human embryonic kidneys at 16 weeks (Ta-
ble S1). Indeed, Li et al. (2016) reported that NGFR was not
applicable to iPSC-derived NPCs, and our preliminary trials
on these three markers for isolation of iPSC-derived NPCs
were also unsuccessful. These results may indicate that arti-
ficially induced NPCs are not completely identical to those
in vivo. Alternatively, iPSC-derived NPCs could represent
those at 4–5 weeks of gestation when NPCs first appear
in vivo (Lindstrom et al., 2018a). Establishment of the
gene expression profiles in human embryonic NPCs at
earlier time points will be helpful toward the identification
of more reliable surface markers for iPSC-derived NPCs.
Nonetheless, our protocol will be useful for a variety of set-
tings in iPSC-based research. Differentiation efficiencies
vary from clone to clone, and can sometimes be very low.
Even in these situations, our protocol allows propagation
of NPCs with high efficiency, and the cells can be subjected
to subsequent functional assays for nephrogenesis in vitro
and in vivo. We recently reported that iPSCs derived from a
patient with a NEPHRIN mutation showed impaired forma-
tion of slit diaphragms in podocytes (Tanigawa et al., 2018).
We also established a selective podocyte induction protocol
from human iPSCs (Yoshimura et al., 2019). Combined with
these two methods, our expansion protocol would be readily
applicable for disease modeling and drug screening using pa-
tient-derived iPSCs. We also reported the generation of
branching ureteric buds from mouse ESCs and human iPSCs
in vitro (Taguchi and Nishinakamura, 2017). When mouse
ESC-derived nephron progenitors and ureteric bud were
combined with mouse embryo-derived stromal cells, we
were able to reconstitute a higher-order kidney structure
in vitro. While we need to induce stromal progenitors from
human iPSCs, our method for expansion of nephron pro-
genitors will be useful for reconstructing a three-dimen-
sional human kidney in the near future.
In summary, we have developed an activin-based proto-
col for maintenance and propagation of iPSC-derived hu-
man NPCs at high efficiency, which can serve as a solid
basis for regenerative medicine of the kidney.
EXPERIMENTAL PROCEDURES
Establishment of SIX2-GFP iPSCs
We established SIX2-GFP iPSCs by transfecting the targeting vec-
tor, along with the TALEN expression plasmids (Sakuma et al.,
2013), into human iPSCs (201B7) (Takahashi et al., 2007). A full
description of the methods is provided in Supplemental Experi-
mental Procedures. We obtained 23 heterozygous GFP-knockin
clones among 91 clones, as determined by PCR, Southern blotting,
and sequencing of the non-GFP-containing allele. Clone #12 was
adapted to the feeder-free condition using StemFit AK02N medium
(Takara Bio, AJ100) and plates coated with iMatrix-511 silk (Nippi,
#892021), as described previously (Nakagawa et al., 2014; Yoshi-
mura et al., 2017), and further electroporated with a plasmid ex-
pressing Cre recombinase to delete the puromycin-resistant
cassette. The resultant colonies were picked up in duplicate and pu-
romycin-sensitive clones were expanded. The absence of the
cassette was verified by PCR, and clones #12-8 and #12-12 were
used for detailed analyses. We used clone #12-12 for most of the
presented data, but consistent data were obtained for clone
#12-8 (see Figure S3).
Propagation of Human iPSC-Derived NPCs
NPC induction was performed based on our previously established
protocol (Taguchi et al., 2014; Taguchi and Nishinakamura, 2017).
Induced NPCs were dissociated by incubation with AccuMAX
(Millipore) for 8 min, and IGTA8+/PGDFRA
/SIX2-GFP+ NPCs
were sorted as described by Kaku et al. (2017). The sorted NPCs
were aggregated as described by Tanigawa et al. (2016) in the
following maintenance medium. The maintenance medium for
NPCs comprised 200 ng/mL recombinant human FGF9 ( R&D Sys-
tems), 10 ng/mL recombinant human TGFa (Peprotech), 1 mg/mL
heparin (Sigma), 10 mM Y27632 (Wako Chemicals), 500 nM
CHIR99021 (Axon Medchem), 5 ng/mL recombinant human LIF
(Millipore), 500 nM DAPT (Millipore), and 50 ng/mL recombinant
human activin A (R&D Systems) in the medium used for NPC
induction from iPSCs (Taguchi et al., 2014). The chemical com-
pounds investigated for NPC culture were as follows: A83-01 (Toc-
ris, #2939); LDN193189 (Wako, #126-05851); SB202190 (Sigma,
#S7067); and SP600125 (Wako, #198-14821). For passaging, the
cells were rinsed with PBS, dissociated with AccuMAX for 8 min,
and split at a 1:3 ratio for passage every 7 days. For cryopreserva-
tion, 7- day cultured cells were rinsed with PBS and dissociated
with AccuMAX for 8 min. After centrifugation, the cells were
washed with medium containing 50 ng/mL DNase I, resuspended
(100,000 cells/vial) in Stem Cell Banker (Nippon Zenyaku Kogyo),
and stored in liquid nitrogen after 24 h of pre-storage at
80 C. De-
frosted cells were incubated overnight in maintenance medium to
allow aggregate formation on low-binding 96-well plates (Tani-
gawa et al., 2016). NPC aggregates were cultured with embryonic
spinal cord for 3 days to initiate differentiation, transferred to
transwell plates, and cultured for an additional 9–20 days in previ-
ously described medium (Li et al., 2016) containing 5% KnockOut
Serum Replacement (Thermo Fisher Scientific).
Establishment of New iPSC Lines
Two iPSC lines (RN7 and RN12) were established from the periph-
eral blood of a healthy volunteer using a previously described
method (Soga et al., 2015). In brief, four reprogramming transcrip-
tion factors (OCT3/4, SOX2, KLF4, and MYC) were introduced us-
ing a Sendai virus vector. After establishing the iPSC clones on
feeder cells (murine embryonic fibroblasts), the temperature-sensi-
tive Sendai virus was eliminated by culturing the cells at 38 C for
24 h. Elimination of the exogenous genes was confirmed by nested
PCR of the vector sequences. RN7 and RN12 cells were adapted to
the feeder-free condition using StemFit AK02N medium and plates
coated with iMatrix-511 silk, as described previously (Nakagawa
et al., 2014; Yoshimura et al., 2017). NPC induction was performed
based on our previously established protocol (Taguchi et al., 2014;
Taguchi and Nishinakamura, 2017).
All animal experiments were carried out in accordance with our
institutional guidelines and were approved by the Ethics Commit-
tee of Kumamoto University (#A29-040). The ethics committees
involved were the Ethics Committee for Epidemiological and Gen-
eral Research at the Faculty of Life Science, Kumamoto University,
and the Ethics Committee for Human Genome and Gene Analysis
Research at the Faculty of Life Science, Kumamoto University
(approval numbers 1453 and 359, respectively).
Statistical Analysis
All experiments were performed independently at least three
times. Data are presented as mean ± SEM. Data were evaluated by
Student’s t test. Differences were considered significant for values
of p < 0.01.
ACCESSION NUMBERS
The accession number for the RNA-seq data deposited with the Na-
tional Center for Biotechnology Information Gene Expression
Omnibus is GEO: GSE118729.
SUPPLEMENTAL INFORMATION
Supplemental Information can be found online at https://doi.org/
10.1016/j.stemcr.2019.07.003.
AUTHOR CONTRIBUTIONS
Culture and analysis of iPSC-derived NPCs, S.T.; generation of
SIX2-GFP iPSCs, Y.K.; transplantation, H.N.; TALEN construction,
T.S. and T.Y.; providing NPC induction protocol, A.T.; writing,
S.T., A.T., and R.N.; Funding acquisition, S.T. and R.N.; Project
administration, R.N.
ACKNOWLEDGMENTS
We thank Tomoko Ohmori, Yumi Soejima, Koichiro Miike, and
Sayoko Fujimura for technical assistance, and Drs. Masaki Hosoya
and Yoshiharu Sato for the RNA-seq analysis. We also thank Alison
Sherwin, PhD, from Edanz Group (www.edanzediting.com/ac) for
editing a draft of the manuscript. This study was supported, in part,
by KAKENHI grants (JP16K19494 to S.T. and JP17H06177 to R.N.)
from the Japan Society for the Promotion of Science, and a grant
from the Research Center Network for Realization of Regenerative
Medicine, Japan Agency for Medical Research and Development.
This work was also supported, in part, by The Mochida Memorial
Foundation for Medical and Pharmaceutical Research.
Received: August 17, 2018
Revised: July 1, 2019
Accepted: July 1, 2019
Published: August 1, 2019
Stem Cell Reports j Vol. 13 j 322–337 j August 13, 2019 335
REFERENCES
Barak, H., Huh, S.H., Chen, S., Jeanpierre, C., Martinovic, J., Pari-
sot, M., Bole-Feysot, C., Nitschke, P., Salomon, R., Antignac, C.,
et al. (2012). FGF9 and FGF20 maintain the stemness of nephron
progenitors in mice and man. Dev. Cell 22, 1191–1207.
Brown, A.C., Adams, D., de Caestecker, M., Yang, X., Friesel, R., and
Oxburgh, L. (2011). FGF/EGF signaling regulates the renewal of
early nephron progenitors during embryonic development. Devel-
opment 138, 5099–5112.
Brown, A.C., Muthukrishnan, S.D., Guay, J.A., Adams, D.C., Scha-
fer, D.A., Fetting, J.L., and Oxburgh, L. (2013). Role for compart-
mentalization in nephron progenitor differentiation. Proc. Natl.
Acad. Sci. U S A 110, 4640–4645.
Brown, A.C., Muthukrishnan, S.D., and Oxburgh, L. (2015). A syn-
thetic niche for nephron progenitor cells. Dev. Cell 34, 229–241.
Chen, S., Brunskill, E.W., Potter, S.S., Dexheimer, P.J., Salomonis,
N., Aronow, B.J., Hong, C.I., Zhang, T., and Kopan, R. (2015).
Intrinsic age-dependent changes and cell-cell contacts regulate
nephron progenitor lifespan. Dev. Cell 35, 49–62.
Costantini, F., and Kopan, R. (2010). Patterning a complex organ:
branching morphogenesis and nephron segmentation in kidney
development. Dev. Cell 18, 698–712.
Dekel, B., Burakova, T., Arditti, F.D., Reich-Zeliger, S., Milstein, O.,
Aviel-Ronen, S., Rechavi, G., Friedman, N., Kaminski, N., Passwell,
J.H., et al. (2003). Human and porcine early kidney precursors as a
new source for transplantation. Nat. Med. 9, 53–60.
Esquela, A.F., and Lee, S.J. (2003). Regulation of metanephric kid-
ney development by growth/differentiation factor 11. Dev. Biol.
257, 356–370.
Fetting, J.L., Guay, J.A., Karolak, M.J., Iozzo, R.V., Adams, D.C.,
Maridas, D.E., Brown, A.C., and Oxburgh, L. (2014). FOXD1 pro-
motes nephron progenitor differentiation by repressing decorin
in the embryonic kidney. Development 141, 17–27.
Harari-Steinberg, O., Metsuyanim, S., Omer, D., Gnatek, Y., Ger-
shon, R., Pri-Chen, S., Ozdemir, D.D., Lerenthal, Y., Noiman, T.,
Ben-Hur, H., et al. (2013). Identification of human nephron pro-
genitors capable of generation of kidney structures and functional
repair of chronic renal disease. EMBO Mol. Med. 5, 1556–1568.
Hartman, H.A., Lai, H.L., and Patterson, L.T. (2007). Cessation of
renal morphogenesis in mice. Dev. Biol. 310, 379–387.
Kaku, Y., Taguchi, A., Tanigawa, S., Haque, F., Sakuma, T., Yama-
moto, T., and Nishinakamura, R. (2017). PAX2 is dispensable for
in vitro nephron formation from human induced pluripotent
stem cells. Sci. Rep. 7, 4554.
Kobayashi, A., Valerius, M.T., Mugford, J.W., Carroll, T.J., Self, M.,
Oliver, G., and McMahon, A.P. (2008). Six2 defines and regulates
a multipotent self-renewing nephron progenitor population
throughout mammalian kidney development. Cell Stem Cell 3,
169–181.
Li, Z., Araoka, T., Wu, J., Liao, H.K., Li, M., Lazo, M., Zhou, B., Sui,
Y., Wu, M.Z., Tamura, I., et al. (2016). 3D culture supports long-
term expansion of mouse and human nephrogenic progenitors.
Cell Stem Cell 19, 516–529.
336 Stem Cell Reports j Vol. 13 j 322–337 j August 13, 2019
Lindstrom, N.O., McMahon, J.A., Guo, J., Tran, T., Guo, Q., Rut-
ledge, E., Parvez, R.K., Saribekyan, G., Schuler, R.E., Liao, C.,
et al. (2018a). Conserved and divergent features of human and
mouse kidney organogenesis. J. Am. Soc. Nephrol. 29, 785–805.
Lindstrom, N.O., Guo, J., Kim, A.D., Tran, T., Guo, Q., De Sena
Brandine, G., Ransick, A., Parvez, R.K., Thornton, M.E., Basking,
L., et al. (2018b). Conserved and divergent features of mesen-
chymal progenitor cell types within the cortical nephrogenic
niche of the human and mouse kidney. J. Am. Soc. Nephrol. 29,
806–824.
Morizane, R., Lam, A.Q., Freedman, B.S., Kishi, S., Valerius, M.T.,
and Bonventre, J.V. (2015). Nephron organoids derived from hu-
man pluripotent stem cells model kidney development and injury.
Nat. Biotechnol. 33, 1193–1200.
Muller, U., Wang, D., Denda, S., Meneses, J.J., Pedersen, R.A., and
Reichardt, L.F. (1997). Integrin alpha8beta1 is critically important
for epithelial-mesenchymal interactions during kidney morpho-
genesis. Cell 88, 603–613.
Muthukrishnan, S.D., Yang, X., Friesel, R., and Oxburgh, L. (2015).
Concurrent BMP7 and FGF9 signalling governs AP-1 function to
promote self-renewal of nephron progenitor cells. Nat. Commun.
6, 10027.
Nakagawa, M., Taniguchi, Y., Senda, S., Takizawa, N., Ichisaka, T.,
Asano, K., Morizane, A., Doi, D., Takahashi, J., Nishizawa, M.,
et al. (2014). A novel efficient feeder-free culture system for the
derivation of human induced pluripotent stem cells. Sci. Rep. 4,
3594.
O’Brien, L.L., Guo, Q., Lee, Y., Tran, T., Benazet, J.D., Whitney, P.H.,
Valouev, A., and McMahon, A.P. (2016). Differential regulation of
mouse and human nephron progenitors by the Six family of tran-
scriptional regulators. Development 143, 595–608.
Park, J.S., Ma, W., O’Brien, L.L., Chung, E., Guo, J.J., Cheng, J.G.,
Valerius, M.T., McMahon, J.A., Wong, W.H., and McMahon, A.P.
(2012). Six2 and Wnt regulate self-renewal and commitment of
nephron progenitors through shared gene regulatory networks.
Dev. Cell 23, 637–651.
Pode-Shakked, N., Gershon, R., Tam, G., Omer, D., Gnatek, Y.,
Kanter, I., Oriel, S., Katz, G., Harari-Steinberg, O., Kalisky, T.,
et al. (2017). Evidence of in vitro preservation of human nephro-
genesis at the single-cell level. Stem Cell Reports 9, 279–291.
Pode-Shakked, N., Pleniceanu, O., Gershon, R., Shukrun, R.,
Kanter, I., Bucris, E., Pode-Shakked, B., Tam, G., Tam, H., Caspi,
R., et al. (2016). Dissecting stages of human kidney development
and tumorigenesis with surface markers affords simple prospective
purification of nephron stem cells. Sci. Rep. 6, 23562.
Rowan, C.J., Li, W., Martirosyan, H., Erwood, S., Hu, D., Kim, Y.K.,
Sheybani-Deloui, S., Mulder, J., Blake, J., Chen, L., et al. (2018).
Hedgehog-GLI signaling in Foxd1-positive stromal cells promotes
murine nephrogenesis via TGFbeta signaling. Development 145.
https://doi.org/10.1242/dev.159947.
Sakuma, T., Ochiai, H., Kaneko, T., Mashimo, T., Tokumasu, D., Sa-
kane, Y., Suzuki, K., Miyamoto, T., Sakamoto, N., Matsuura, S., et al.
(2013). Repeating pattern of non-RVD variations in DNA-binding
modules enhances TALEN activity. Sci. Rep. 3, 3379.
Sharmin, S., Taguchi, A., Kaku, Y., Yoshimura, Y., Ohmori, T., Sa-
kuma, T., Mukoyama, M., Yamamoto, T., Kurihara, H., and Nishi-
nakamura, R. (2016). Human induced pluripotent stem cell-
derived podocytes mature into vascularized glomeruli upon exper-
imental transplantation. J. Am. Soc. Nephrol. 27, 1778–1791.
Soga, M., Ishitsuka, Y., Hamasaki, M., Yoneda, K., Furuya, H., Mat-
suo, M., Ihn, H., Fusaki, N., Nakamura, K., Nakagata, N., et al.
(2015). HPGCD outperforms HPBCD as a potential treatment for
Niemann-Pick disease type C during disease modeling with iPS
cells. Stem Cells 33, 1075–1088.
Takahashi, K., Tanabe, K., Ohnuki, M., Narita, M., Ichisaka, T., To-
moda, K., and Yamanaka, S. (2007). Induction of pluripotent stem
cells from adult human fibroblasts by defined factors. Cell 131,
861–872.
Taguchi, A., Kaku, Y., Ohmori, T., Sharmin, S., Ogawa, M., Sasaki,
H., and Nishinakamura, R. (2014). Redefining the in vivo origin
of metanephric nephron progenitors enables generation of com-
plex kidney structures from pluripotent stem cells. Cell Stem Cell
14, 53–67.
Taguchi, A., and Nishinakamura, R. (2017). Higher-order kidney
organogenesis from pluripotent stem cells. Cell Stem Cell 21,
730–746.
Takasato, M., Er, P.X., Chiu, H.S., Maier, B., Baillie, G.J., Ferguson,
C., Parton, R.G., Wolvetang, E.J., Roost, M.S., Lopes, S.M., et al.
(2016). Kidney organoids from human iPS cells contain multiple
lineages and model human nephrogenesis. Nature 536, 238.
Tanigawa, S., Islam, M., Sharmin, S., Naganuma, H., Yoshimura, Y.,
Haque, F., Era, T., Nakazato, H., Nakanishi, K., Sakuma, T., et al.
(2018). Organoids from nephrotic disease-derived iPSCs identify
impaired NEPHRIN localization and slit diaphragm formation in
kidney podocytes. Stem Cell Reports 11, 727–740.
Tanigawa, S., Sharma, N., Hall, M.D., Nishinakamura, R., and Per-
antoni, A.O. (2015). Preferential propagation of competent SIX2+
nephronic progenitors by LIF/ROCKi treatment of the meta-
nephric mesenchyme. Stem Cell Reports 5, 435–447.
Tanigawa, S., Taguchi, A., Sharma, N., Perantoni, A.O., and Nishi-
nakamura, R. (2016). Selective in vitro propagation of nephron
progenitors derived from embryos and pluripotent stem cells.
Cell Rep. 15, 801–813.
Yoshimura, Y., Taguchi, A., and Nishinakamura, R. (2017). Gener-
ation of a three-dimensional kidney structure from pluripotent
stem cells. Methods Mol. Biol. 1597, 179–193.
Yoshimura, Y., Taguchi, A., Tanigawa, S., Yatsuda, J., Kamba, T., Ta-
kahashi, S., Kurihara, H., Mukoyama, M., and Nishinakamura, R.
(2019). Manipulation of nephron-patterning signals enables selec-
tive induction of podocytes from human pluripotent stem cells.
J. Am. Soc. Nephrol. 30, 304–321.
Stem Cell Reports j Vol. 13 j 322–337 j August 13, 2019 337